rumdl 0.3.2

A fast Markdown linter and formatter written in Rust
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189

use super::*;

fn signature(scan: HtmlCommentScan) -> (Vec<(usize, usize)>, Option<usize>) {
    (
        scan.ranges.into_iter().map(|range| (range.start, range.end)).collect(),
        scan.unterminated,
    )
}

#[test]
fn comment_exclusion_index_preserves_nested_ranges_and_half_open_edges() {
    // These examples separately reject raw-end lookup, an exclusive start,
    // and an inclusive end in the exclusion index.
    let nested = format!("{}<!--x-->", " ".repeat(26));
    assert!(
        scan_html_comments(&nested, &[(3, 40), (20, 25)], &[], 0)
            .ranges
            .is_empty()
    );
    assert!(scan_html_comments("<!--x-->", &[(0, 1)], &[], 0).ranges.is_empty());
    assert_eq!(scan_html_comments("<!--x-->", &[(0, 0)], &[], 0).ranges.len(), 1);
}

#[test]
fn sorted_comment_line_containment_preserves_empty_and_reversed_queries() {
    let ranges = [
        ByteRange { start: 3, end: 8 },
        ByteRange { start: 8, end: 8 },
        ByteRange { start: 10, end: 16 },
    ];
    for start in 0..20 {
        for end in 0..24 {
            assert_eq!(
                is_line_entirely_in_sorted_html_comment(&ranges, start, end),
                is_line_entirely_in_html_comment(&ranges, start, end),
                "start={start}, end={end}",
            );
        }
    }
    assert!(!is_line_entirely_in_sorted_html_comment(&[], 0, 0));
}

#[test]
fn indexed_comment_scan_preserves_range_and_delimiter_boundaries() {
    for content in [
        "",
        "é日本🙂 <!--> <!---> <!-- real --> <!-- unclosed",
        "`<!--` <!-- real --> `-->` <!-- later -->",
        "<!-- open `-->` real --> <!-- end",
        "<!-- <!-- --> --> <!-- -->",
        "---\r\nx: '<!--'\r\n---\r\n<!-- body -->",
    ] {
        let len = content.len();
        let cases = [
            vec![],
            vec![(0, len)],
            vec![(len, len), (usize::MAX, 0), (2, 1)],
            vec![(8, 15), (0, 3), (2, 10), (10, 10)],
            vec![(20, 25), (0, 20), (5, 8), (3, 30)],
            vec![(0, usize::MAX)],
        ];
        let starts: Vec<_> = content
            .char_indices()
            .map(|(byte, _)| byte)
            .chain([len, len + 1, usize::MAX])
            .collect();
        for spans in &cases {
            for blocks in &cases {
                for &start in &starts {
                    assert_eq!(
                        signature(scan_html_comments(content, spans, blocks, start)),
                        signature(legacy_scan_html_comments(content, spans, blocks, start)),
                        "{content:?}, spans={spans:?}, blocks={blocks:?}, start={start}",
                    );
                }
            }
        }
    }
}

proptest::proptest! {
    #![proptest_config(proptest::test_runner::Config::with_cases(256))]
    #[test]
    fn indexed_comment_scan_matches_scalar_with_arbitrary_ranges(
        tokens in proptest::collection::vec(
            proptest::sample::select(vec!["<!--", "-->", "<!-->", "<!--->", "`", "``", "\n", "\r\n", "é日🙂", "text", "<", "!", "-", " "]),
            0..100,
        ),
        span_specs in proptest::collection::vec((0usize..2000, 0usize..40, proptest::bool::ANY), 0..40),
        block_specs in proptest::collection::vec((0usize..2000, 0usize..40, proptest::bool::ANY), 0..40),
        start_choice in 0usize..2000,
    ) {
        let content = tokens.concat();
        let convert = |specs: Vec<(usize, usize, bool)>| -> Vec<(usize, usize)> {
            specs.into_iter().map(|(offset, length, reversed)| {
                let start = offset % (content.len() + 16);
                let end = start + length;
                if reversed { (end, start) } else { (start, end) }
            }).collect()
        };
        let spans = convert(span_specs);
        let blocks = convert(block_specs);
        let boundaries: Vec<_> = content.char_indices().map(|(byte, _)| byte).chain([content.len(), content.len() + 1]).collect();
        let start = boundaries[start_choice % boundaries.len()];
        proptest::prop_assert_eq!(
            signature(scan_html_comments(&content, &spans, &blocks, start)),
            signature(legacy_scan_html_comments(&content, &spans, &blocks, start)),
        );
    }

    #[test]
    fn sorted_comment_line_containment_matches_scalar(
        specs in proptest::collection::vec((0usize..10, 0usize..80), 0..60),
        start in 0usize..4000,
        width in 0usize..100,
        reverse in proptest::bool::ANY,
    ) {
        let mut next = 0;
        let ranges: Vec<_> = specs.into_iter().map(|(gap, len)| {
            let start = next + gap;
            let end = start + len;
            next = end;
            ByteRange { start, end }
        }).collect();
        let end = if reverse { start.saturating_sub(width) } else { start + width };
        proptest::prop_assert_eq!(
            is_line_entirely_in_sorted_html_comment(&ranges, start, end),
            is_line_entirely_in_html_comment(&ranges, start, end),
        );
    }
}

// Frozen scalar implementation, including delimiter and unterminated-comment behavior.
fn legacy_scan_html_comments(
    content: &str,
    code_span_ranges: &[(usize, usize)],
    code_block_ranges: &[(usize, usize)],
    scan_from: usize,
) -> HtmlCommentScan {
    let in_code = |pos: usize| {
        code_span_ranges.iter().any(|&(start, end)| pos >= start && pos < end)
            || code_block_ranges.iter().any(|&(start, end)| pos >= start && pos < end)
    };

    let mut ranges = Vec::new();
    let mut search_from = scan_from.min(content.len());
    while let Some(rel) = crate::utils::find_ascii_delimiter(&content[search_from..], "<!--") {
        let open = search_from + rel;
        if in_code(open) {
            // Literal `<!--` inside code: not a comment opener.
            search_from = open + "<!--".len();
            continue;
        }
        // Find the next `-->` that is not itself inside code. The search starts
        // two bytes into the opener, because the opener's own trailing dashes can
        // serve as the closer's leading dashes: `<!-->` and `<!--->` are complete
        // comments, and a line holding one ends an HTML block.
        let mut close_from = open + 2;
        let end = loop {
            let Some(crel) = crate::utils::find_ascii_delimiter(&content[close_from..], "-->") else {
                break None;
            };
            let close = close_from + crel;
            if in_code(close) {
                close_from = close + "-->".len();
                continue;
            }
            break Some(close + "-->".len());
        };
        match end {
            Some(end) => {
                ranges.push(ByteRange { start: open, end });
                search_from = end;
            }
            // Unterminated comment (no closer anywhere): it covers no range,
            // and no opener after it can close either, so the scan is done.
            None => {
                return HtmlCommentScan {
                    ranges,
                    unterminated: Some(open),
                };
            }
        }
    }
    HtmlCommentScan {
        ranges,
        unterminated: None,
    }
}